Project Summary Brain imaging genetics studies the relationship between genetic variations and brain imaging quantitative traits (QTs) and offers enormous potential to reveal the genetic underpinning of the neurobiological system that can impact the development of diagnostic, therapeutic and preventative approaches for complex brain disorders. Two critical gaps limiting the progress of brain imaging genetics include (1) the unprecedented scale and complexity of the imaging genetic data sets, and (2) lack of intermediate-level omics data to capture the molecular effects linking genetics to brain QTs. Our prior studies have contributed substantially to addressing the first gap. The proposed project will develop new informatics strategies to bridge the second gap, where valuable existing data in the omics domain will be leveraged to link brain imaging and genetics. In this project, we will focus on transcriptomics, and will make use of major transcriptomics data repositories including Genotype-Tissue Expression (GTEx) Project, UK Brain Expression Consortium (UKBEC), and Allen Human Brain Atlas (AHBA). Our overarching goal is to identify brain imaging genetic associations with evidence manifested in the human brain transcriptome. Our hypothesis is that, with additional source of evidence at the transcriptomic level, the identified brain imaging genetic associations are biologically more meaningful and less likely to be false positives. To achieve our goal, we propose four aims. Aim 1 is to develop novel bi-multivariate models incorporating regional tissue-specific expression quantitative trait locus (eQTL) knowledge for mining brain imaging genetic associations. Given that eQTL is a source of tissue-specific evidence to link genotype, gene expression, and brain QTs, we will develop novel eQTL-guided bi-multivariate models to identify imaging genetic associations potentially evidenced by regional tissue-specific eQTL knowledge. Aim 2 is to develop novel bi-multivariate models incorporating brain-wide genome-wide (BWGW) cross-domain co-expression patterns for mining brain imaging genetics associations. AHBA, a BWGW gene expression database, is a natural connection between genome and brain. We propose to develop novel biclustering and bi-multivariate methods to identify meaningful AHBA modules with cross-domain co-expression patterns, and use these patterns to guide the search for co-expression-aware associations between genetic variations and multimodal brain imaging measures. Aim 3 is to develop open source software tools for structure-aware mining of brain imaging genetic associations. Aim 4 is to perform evaluation and validation on both simulated data and real imaging genetics cohorts. Successful completion of the above aims will produce innovative informatics methods and tools for integrative analysis of imaging, genetics and transcriptomics data to address a critical barrier in brain imaging genetics. Using ADNI and related cohorts as test beds, these methods and tools will be shown to have considerable potential for understanding the molecular mechanism of Alzheimer?s disease, and be expected to impact neurological and psychiatric research in general and benefit public health outcomes.